INDEX

Explanations

security packaging and features

The main thing this neuron does is detect mentions of protective packaging materials (e.g. anti-static bags, foam, shrink-wrap tubing, etc.).

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

see

-1.34

 tungkol

-1.28

 lapin

-1.25



-1.24

艷

-1.23

 hansen

-1.16

 douglas

-1.16

kya

-1.15

 only

-1.13

 davis

-1.11

POSITIVE LOGITS

 jeweiligen

1.30

或者是

1.28

スコット

1.27

expiresIn

1.22

 bereits

1.20

小结

1.19

约定

1.17

 вже

1.16

 pretrial

1.16

而非

1.16

Activations Density 0.108%